config.yaml

---


# Primers - uncomment the correct set
# ====================================
# These are some common ones used at our institution.  You may need to provide your own
# primer sequences

# Illumina V3-V4 protocol primers
fwd_primer: "TCGTCGGCAGCGTCAGATGTGTATAAGAGACAGCCTACGGGNGGCWGCAG"
rev_primer: "GTCTCGTGGGCTCGGAGATGTGTATAAGAGACAGGACTACHVGGGTATCTAATCC"
fwd_primer_rc: "CTGCWGCCNCCCGTAGGCTGTCTCTTATACACATCTGACGCTGCCGACGA"
rev_primer_rc: "GGATTAGATACCCBDGTAGTCCTGTCTCTTATACACATCTCCGAGCCCACGAGAC"


# # Surette V3 protocol
#fwd_primer: "ATTACCGCGGCTGCTGG"
#rev_primer: "CCTACGGGAGGCAGCAG"
#fwd_primer_rc: "CCAGCAGCCGCGGTAAT"
#rev_primer_rc: "CTGCTGCCTCCCGTAGG"

# Surette V3-V4 protocol
# fwd_primer: "NNNNGGACTACHVGGGTWTCTAAT"
# rev_primer: "CCTACGGGAGGCAGCAG"
# fwd_primer_rc: "ATTAGAWACCCBDGTAGTCCNNNN"
# rev_primer_rc: "CTGCTGCCTCCCGTAGG"

# Harrison lab protocol
#fwd_primer: "CCTACGGGNGGCWGCAG"
#rev_primer: "GACTACHVGGGTATCTAATCC"
#fwd_primer_rc: "CTGCWGCCNCCCGTAGG"
#rev_primer_rc: "GGATTAGATACCCBDGTAGTC"


run_name: "test" 

# inital quality trim, trim the 3' end of the reads at the given quality cutoff
# best to keep this value low if you can to keep the 3' ends long for overlapping
# # but if you have really bad 3' qualities this can help improve the read pair merging
q_trim: 20

# max value for expected error for read filtering. Increase to reduce stringency.
max_expected_error: 1.0

# Truncate all reads at this length
read_length_cutoff: 400

# number of threads for multi-threaded applications 
# overriden by -j option to snakemake when running multithreaded
# number of threads per-node when submitting to cluster
num_threads: 20

# Directory where the raw reads are stored
read_directory: "reads"

# database for dada2::assignTaxonomy
database: "~/databases/assignTaxonomy/rdp_train_set_14.fa.gz"

# one of "rdp", "gg", "silva"
taxa_type: "rdp"

# for the Surette lab protocol, or any other where the forward 16S primer is on the reverse read
# in order for the taxonomy assignment to work correctly the reads need to be in the same direction
# as the database.  For most protocols this should be FALSE.  If you aren't getting any taxonomy 
# assigned to your reads try changing this to TRUE and see if that makes a difference.
revcomp: "FALSE"

# Unique sample descriptors - should be unique, present in the file names and contain no underscores.  The 
# current standard for fastq filenaming from Illumina is to put the sample name at the beginning of the 
# file followed by an underscore.  This is the convention the pipeline follows.  If your files are named 
# differently then you can modify the input step in the clip_primers rule in the snakefile. 
samples: ["S1", "S2", "S3", "S4"]